Contents I. Introduction
In REAL-WORLD inverse problems, the objective function is generally characterized by a multimodal function. Since traditional optimal methods are not very suitable for finding the global optimal solution of these kinds of problems, stochastic algorithms such as genetic, simulated annealing, tabu search, and ant colony algorithms have recently been developed. Noting, however, that thousands of iterations are generally required in stochastic algorithms, the aforementioned algorithms are not particularly suitable for inverse problems that require heavy computation, and hence many efforts must be put in before they can be used to address real-life engineering problems.
